USDA Commodity Costs and Returns (CAR) and Monthly Milk Cost-of-Production (COP): A 2016 Data Product Review and Proposals for Change, Staff Paper P16-4

Recommendations
Based on the Data Product review, the panel recommends the following actions:
Near-term
• Add the costs and returns reports to the Crop Production Practices portion of the ARMS data tool. Set it up to allow viewing of multiple columns on the screen at one time if possible.
• Drop the monthly milk cost-of-production estimates.
• For the between-survey years, smooth the estimates retroactively to minimize the jarring jumps that sometimes occur. Use an approach that incorporates the annual price indices, in effect smoothing the quantities.
• Review the corn silage pricing in the regional dairy estimates to see if 1) the surveyed market prices are representative of all dairy farms and 2) they are in line with comparable corn grain prices and differences in harvesting and storage costs of silage and grain.
Longer-term
• Redouble efforts to obtain feedback from stakeholders such as producers and private consultants in addition to the academic perspective represented on this committee.
• If the monthly milk cost-of-production estimates are continued, refine the indexing method to reflect the cost shares of a more detailed list of feeds, and possibly do separate estimates for the higher-cost region (CA and the northeast) and the lower-cost central part of the U.S.
• Review the estimation procedure for unpaid labor, including procedures for computing both the quantity and price of unpaid labor.
• Compare the CAR machinery costs with information available from custom rate surveys currently done by state extension staff, and with a NASS national custom rate survey if that effort moves to completion.
• Review the information generated by the major university farm record systems on a regular basis as a way of identifying possible improvements in the CAR methodology. Also, investigate ways of obtaining feedback on-line from the data user community.
• Investigate alternative data sources and methods to ensure that the spatial variability of input prices is reflected in the estimates whenever possible.
• Enhance the documentation of the CAR data by adding material including a discussion of economic cost concepts, a more detailed description of how CAR estimates are set between surveys, links to the commodity costs and return questionnaires, and measures of the variability of the estimates such as coefficients of variation.